Teen tobacco use has been rising steadily since the early 1990's after a decade of relative stability. Compared with teens in traditional high schools, smoking rates for teens in alternative high schools are alarmingly high; approximately 80 percent have ever-smoked cigarettes and of these, 70 percent are experimental or light smokers. Because schools are the most efficient way to reach teens, it is critical to find effective school-based strategies that reach teens who are at the highest risk of becoming addicted to tobacco. Using a randomized trial, we propose to develop and evaluate a new tobacco prevention curriculum (Youth Advocacy Program) for very high-risk, ethnically diverse 11th and 12th grade students attending 10 alternative high schools in the San Jose area of California. The Youth Advocacy Program curriculum (treatment) will be tested against a placebo-attention substance abuse curriculum (control). The intervention offers the first empirical test of an intervention that seeks to prevent long-term addiction to tobacco by teens who have high rates of experimental smoking by: 1) modifying proximal social influences on smoking, e.g., perceived norms and values about smoking; 2) building awareness of distal environmental influences on smoking that stimulate teen tobacco use by increasing acceptability and availability of cigarette use in the community, and 3) engaging youth in devising and implementing strategies to modify pro-smoking community influences. Our aim is to reverse the usual progression from experimental smoking to regular smoking by teens in the Youth Advocacy Program. Our primary hypothesis is that experimental/light smokers in the treatment schools will experience a 35 percent quit rate at the end of the semester-long intervention compared with a 5 percent quit rate among experimental smokers in the control schools, and that the reduction in smoking will be maintained for 6 months following the intervention. Twenty-five students will be recruited from each of the 10 alternative schools. Five schools will be randomly assigned to the Youth Advocacy Program curriculum and five to the placebo-attention curriculum. The intervention will be repeated for 4 semesters (2 years), resulting in 500 treatment and 500 control students. The intervention will be intensive and will include 25 hours of classroom activities, 15 hours of a weekend Youth Advocacy Institute, and 20 hours of community-based advocacy activities. To assess change, teens at the treatment and control schools will complete three surveys; at pre- intervention, post-intervention, and 6 months post-intervention. We will test our primary hypothesis using a logistic regression analysis, with current smoking (yes/no) as the dependent measure and treatment (T/C), school-within-treatment, and semester-within-school-within-treatment as the effects of interest. The results will be confirmed using a logistic regression analysis that incorporates a fixed-within-analysis correlation (intraclass correlation).